Development of a Heterogeneity Analysis Framework for Collaborative Sponge City Management

Bing Li, Shoulong Dong, Yue Huang, Guang WangWater 2019 · DOI

TL;DR

This paper introduces a novel framework using Principal Component Analysis (PCA) and Self-Organizing Feature Maps (SOM) to analyze the heterogeneity of Chinese cities for "Sponge City" construction. By clustering cities based on diverse environmental and socio-economic factors, the study reveals regional patterns and lays the groundwork for more effective, collaborative urban water management strategies.

What Problem Does This Paper Address?

Rapid urbanization and climate change are exacerbating urban water challenges like flooding, water scarcity, and pollution globally. The "Sponge City" concept, aiming to manage rainwater more sustainably, has been adopted in China. However, a significant gap exists in understanding how different cities' unique characteristics (heterogeneity) impact sponge city implementation and how to foster collaborative management between cities. Current approaches often rely on generic guidelines, overlooking the diverse local conditions that can lead to the failure of imported strategies. This paper aims to bridge this gap by providing a systematic way to analyze and understand urban heterogeneity for better sponge city planning and collaboration.

Methods

The core of this research is a two-step analytical framework:

  1. Principal Component Analysis (PCA): PCA is used to reduce the dimensionality of a large set of indicators (102 in this study) related to natural environment, infrastructure, water pollutant, social development, water resources, and precipitation. This process identifies the principal components that explain the most variance, effectively capturing the underlying reasons for heterogeneity among cities. It also helps in standardizing and combining indicators that might have different units of measurement.
  2. Self-Organizing Feature Map (SOM): A SOM, an unsupervised artificial neural network, is then employed to cluster the cities based on the principal components derived from PCA. SOMs are chosen for their ability to create a low-dimensional, topologically ordered representation of the input data, allowing for visual identification of city groups with similar characteristics. The study used a 4x4 SOM topology, resulting in 16 clusters.

This combination of PCA for feature extraction and SOM for clustering is a standard approach in data analysis, but its application to the specific problem of sponge city heterogeneity analysis is novel.

Datasets

The study utilized data from 96 Chinese cities. The data was sourced from:

  • 2017 China Statistical Yearbook
  • China Urban Construction Statistical Yearbook
  • 2017 China City Statistical Yearbook
  • China Water Statistics Yearbook 2017
  • China Meteorological Data Network

The paper selected 102 sub-indicators across various categories, with precipitation data including monthly averages and maximum daily rainfall, and temporal differences described by 12 variables (January to December).

Key Insights and Ideas

  • Regional Clustering is Evident: Cities with similar geographical locations and meteorological conditions tend to cluster together, suggesting that regional environmental factors are strong drivers of sponge city characteristics.
  • Heterogeneity is Multi-faceted: While geography plays a role, socio-economic factors and infrastructure development also significantly contribute to urban heterogeneity, leading to cities in the same region being placed in different clusters (e.g., Zhuhai and Guangzhou).
  • PCA Reveals Key Drivers: PCA identified that precipitation characteristics and developed urban area are major factors differentiating cities in the southeast from the northwest. Urban afforestation and temperature were found to be key differentiators for southeastern cities like Shanghai.
  • Potential for Interregional Collaboration: The framework demonstrates that even cities with distant geographical spaces but similar underlying characteristics can be grouped, highlighting the potential for interregional collaborative sponge city strategies.
  • Data Limitations are a Hurdle: The study acknowledges that data limitations, particularly regarding construction and maintenance costs, hinder stakeholders' willingness to invest in sponge city solutions and quantitative evaluations.

Key Results

  • The study analyzed 96 Chinese cities using 102 sub-indicators.
  • PCA identified principal components that explained significant variance: the first two explained 58%, the first six explained 75%, and the first ten explained 85% of the total variations.
  • A 4x4 SOM topology resulted in 16 distinct clusters of cities.
  • The clustering revealed that cities in the same geographical region often share similar meteorological conditions and thus can be grouped together (e.g., northwestern cities in Neuron 5, southeastern cities in Neuron 15).
  • However, the study also showed that cities within the same province can be in different clusters (e.g., Zhuhai and Guangzhou), indicating that extreme precipitation events and socio-economic factors can override geographical proximity.
  • Similarly, Shanghai and Nanjing were separated due to differences in afforestation ratio and road network density.

Limitations and Future Work

Limitations:

  • Data Availability and Reliability: The study acknowledges that some crucial indicators were screened out due to unavailability or low reliability for a large number of cities.
  • Cost Data: Regional data on construction and maintenance costs for sponge city implementation is limited, which impacts stakeholder willingness to invest and hinders quantitative evaluation.
  • Scope: While the study used 96 major Chinese cities, the authors suggest that urban water challenges are similar in many developing countries, implying the need for broader application.

Future Work:

  • Advanced Data Management: The authors suggest using advanced modeling and data management techniques like Bayesian networks and Monte Carlo simulations to address data uncertainty and improve sponge city concept failure risk assessment.
  • Top-Level, Regional Collaboration: Future strategies should focus on regional collaborative management, potentially disregarding administrative boundaries and prioritizing cities with similar ecological conditions and adjacent geographical spaces.
  • Interdepartmental and Stakeholder Cooperation: Strengthening cooperation and linkage between different administrative departments and stakeholders is crucial for design, construction, and management.
  • Integration of Technologies: Incorporating artificial intelligence and the Internet of Things (IoT) is recommended to improve the overall efficiency of sponge city construction and management.

Summary updated

Abstract

Rapid urbanization, inappropriate urban planning and the changing climate in many countries have resulted in flooding, water shortage and water pollution around the world. Although the sponge city concept has been applied in both macro-scales and micro-scales to address those challenges, research on the heterogeneity of different cities for sponge city construction and the collaborative management between cities is insufficient. Therefore, this paper proposes a multivariate cluster analysis framework and conducts an empirical study using 96 Chinese cities. By considering the local infrastructure, economic development, water resource distribution, water quality and precipitation characteristics in each city, and integrating the principal component analysis and a self-organizing feature mapping network, this paper shows the potential of regional and interregional sponge city collaborative management. This will provide an opportunity for developing a new sponge city management mechanism and will promote the establishment of multi-functional departments for urban flood control and water quality improvement.
BibTeX
@inproceedings{li2019development,
  author = {Bing Li and Shoulong Dong and Yue Huang and Guang Wang},
  title = {{Development of a Heterogeneity Analysis Framework for Collaborative Sponge City Management}},
  booktitle = {Water},
  year = {2019},
  doi = {10.3390/w11101995},
}
Figure